[pymvpa] Searchligh SVM on fMRI Beta Values in PyMVPA

Raúl Hernández raul at lafuentelab.org
Thu May 23 21:55:02 BST 2019


For sure there is a better approach, but I also had the same issue. What I
ended up doing is pretending that each beta map was a volume of a raw fMRI
acquisition. So I calculated the beta maps and then used the FSL command
fslmerge to concatenate every beta map to analyze. Then I put the resulting
4D in a folder following the structure required for PyMVPA and loaded the
file as if it were a raw fMRI. Then proceeded with my searchlight script
just skipping the preprocessing.



On Thu, May 23, 2019 at 10:09 PM Alireza Karami <alireza.kr at gmail.com>

> Hi
> I have a dataset of an fMRI experiment and I have preprocessed the data
> and done first level analysis. Now, I want to do searchlight SVM on the *beta
> values* like the demo fmri searchlight lda
> <http://www.cosmomvpa.org/matlab/demo_fmri_searchlight_lda.html> [1] in
> CoSMoMVPA. I want to do this with PyMVPA. I know how to do searchlight SVM
> on raw fMRI data in PyMVPA
> <http://www.pymvpa.org/examples/searchlight.html> [2] but I do not know
> how to import beta values as a dataset in PyMVPA and do searchlight SVM on
> them.
> Would someone help me to do this?
> [1] http://www.cosmomvpa.org/matlab/demo_fmri_searchlight_lda.html
> [2] http://www.pymvpa.org/examples/searchlight.html
> Thanks
> _______________________________________________
> Pkg-ExpPsy-PyMVPA mailing list
> Pkg-ExpPsy-PyMVPA at alioth-lists.debian.net
> https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/pkg-exppsy-pymvpa
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://alioth-lists.debian.net/pipermail/pkg-exppsy-pymvpa/attachments/20190523/338a65f8/attachment.html>

More information about the Pkg-ExpPsy-PyMVPA mailing list